58 WALTER S. HUNTER AND JOS. U. YARBROUGH 



detail. It will also be recalled that no one of these rats im- 

 proved during his 30-days' training upon B. 



Three untrained rats, Nos. 26, 27 and 29, formed the original 

 h. c. habit as indicated in table 1. They were then trained on 

 B until it was mastered. (I shall discuss certain details of this 

 training in a following section.) At the close of the 4 days 

 on B made at 87|%, these rats were retested on h. c. The 

 results for all save the original learning are given in table 5. 



No 26 required 280 trials for the re-learning here in question. 

 No. 27 required 310; and No. 29, 260 trials. The intervals for 26 

 and 29 are a little too small inasmuch as these two rats grew 

 sick and died. Each, however, had reached 80% correct and 

 so was within 7|% of the standard. The indications, from this 

 test are that marked progress must be made in the formation 

 of a second contradictory habit before the retention of a first 

 habit is noticeably affected. This can be represented graph- 



